Site assessment is a key aspect of understanding the risks of and planning remediafion of hazardous waste sites. Budget constraints, site heterogeneity and uncertainties in assessing exposure complicate assessment. The ideal approach also varies according to the purpose. From a regulatory perspective, goals include general characterization of a site, estimating mean levels for risk assessment, planning remediation, and assessing compliance. In contrast, environmental epidemiologists are interested in estimating individual exposures as accurately as possible to study associations with health outcomes. The objective of this project is to provide statistical design and analysis tools to improve the accuracy and reliability of site and exposure assessment for Superfund hazardous waste sites. Our approach is based on statistical modeling, along with optimal design considerations that maximize prediction accuracy while minimizing cost and accounting for practical considerations. Building on the basic spatial kriging model, our spatial model-based approach to design and analysis will be compared to existing design-based approaches that do not account for spatial correlation. We will extend these to complicated real-worid settings, including the use of previous targeted samples and non-detect, proxy and composite samples, which may allow a reduction in sampling costs. At the megasite scale, we will relate soil concentrations to biomarker levels in the Tar Creek Superfund site, and develop spatial measurement error models for relating environmental concentrations to exposure as measured by biomarkers? thus, accounting for incomplete environmental sampling. We will clarify under what circumstances a spatial model-based approach provides real benefits in pracfice, reducing cost and uncertainty. To make spatial-model based methods accessible to the site-assessment community, we will develop software tools for use by EPA and site professionals.